Seminar New Trends in Machine Learning und Data Analytics
Überblick
Das Seminar ist auf ca. 40 Teilnehmer/innen des Masterstudiengangs Data Science begrenzt.
Um an dem Seminar teilzunehmen, müssen Sie sich im AlmaWeb für das Modul Skalierbare Datenbanktechnologien 1 (10-INF-DS01) und das Seminar anmelden sowie an der Einführungsveranstaltung teilnehmen, bei der die endgültige Platz- und Themenvergabe erfolgt.
** Bei Fragen und Problemen zur An- und Abmeldung wenden Sie sich bitte an das Studienbüro via einschreibung(at)math.uni-leipzig.deDie studentischen Vorträge finden an den Freitagsterminen im Jan./Feb 2022 (jeweils ab 13:15 Uhr) statt.
Alle Materialien und Informationen finden Sie im entsprechenden Moodle-Kurs.
Leistungsbewertung
Ein erfolgreiches Seminar setzt die Teilnahme an allen Seminarterminen voraus, die selbständige Erarbeitung eines Themas sowie einen Vortrag sowie eine schriftliche Ausarbeitung (15-20 Seiten) über das Thema. Die Benotung setzt sich aus der Note zu Vortrag und Diskussion sowie der Note für die Ausarbeitung zusammen. Einige Hinweise zum Verfassen der schriftlichen Ausarbeitung finden Sie hier.
Literatur
Für das Verständnis ist es hilfreich sich folgende Studentenarbeiten oder Originalliteratur Quellen zum Thema Deep Learning anzuschauen.
Thema | Quelle |
---|---|
Einleitung Deep-Learning | Studentenarbeit,[1],[2] |
Autoencoder und CNN | Studentenarbeit,[1] |
Recurrent Neural Networks | [1] [2] |
Themen und Betreuer
Nr | Thema | Betreuer | Votragender | Quellen | Termin Vortrag | Folien | Ausarbeitung |
---|---|---|---|---|---|---|---|
0 | Einführung in Seminar und Themen | - | Prof. Rahm | 22.10.2021 15:15 Uhr | - | ||
Privacy & Security | |||||||
P1 | Model inversion Attack | Schneider | [1] | ||||
P2 | Bypass Facial ID Authentication | Rohde | [1] | ||||
P3 | Private Next-Location Prediction | Schneider | [1] | ||||
P4 | Fair ML | Saaedi | [1] | ||||
P5 | Runtime Analysis of Whole-System Provenance | Grimmer | [1] | ||||
P6 | Unicorn: Runtime provenance-based detector for advanced persistent threats | Grimmer | [1] | ||||
P7 | Cross-Layer Deanonymization Methods in the Lightning Protocol | Kreusch | [1] | ||||
P8 | Blockchain-Federated-Learning for COVID-19 detection using CT-Images | Kreusch | [1] | ||||
Techniques for limited labeled data & heterogeneous data | |||||||
LD1 | Deep Entity Matching with Pre-Trained Language Models | Köpcke | [1] | ||||
LD2 | End-to-end Task Based Parallelization for Entity Resolution on Dynamic Data | Köpcke | [1] | ||||
LD3 | Deep Learning for Blocking in Entity Matching | Franke | [1] | ||||
LD4 | Multi-modal Representation Learning | Wilke | [1],[2],[3],[4] | ||||
LD5 | Self Supervised Learning | Wilke | [1],[2] | ||||
LD6 | On the Ambiguity of Rank-Based Evaluation of Entity Alignment or Link Prediction Methods | Obraczka | [1] | ||||
LD7 | DGL-KE: Training Knowledge Graph Embeddings at Scale | Obraczka | [1] | ||||
Natural Language Processing | |||||||
NLP1 | Pre-trained Models for Natural Language Processing: A Survey | Christen | [1] | ||||
NLP2 | On the Dangers of Stochastic parrots:Can Language Models Be Too Big? | Pogany | [1] | ||||
Evolving Graphs | |||||||
G1 | Programming Abstractions for Distributed Graph Processing | Adameit | [1] | ||||
G2 | Temporal Graph Query Languages | Rost | [1] | ||||
G3 | Tegra: Efficient Ad-Hoc Analytics on Evolving Graphs | Gomez | [1] | ||||
G4 | Effective partitioning mechanisms for time-evolving graphs in the Flink system | Gomez | [1] | ||||
G5 | Graph Stream Summarization Techniques | Rost | [1],[2] | ||||
Graph Machine Learning | |||||||
G6 | GraphFL: A Federated Learning Framework for Semi-Supervised Node Classification on Graphs | Hofer | [1] | ||||
G7 | Stochastic Graph Embedding of Temporal Graphs | Adameit | [1] | ||||
G8 | Attention Temporal Graph Convolutional Networks | Schuchart | [1] | ||||
G9 | Convolutional 2D Knowledge Graph Embeddings | Hofer | [1] | ||||
G10 | Evolving Graph Convolutional Networks | Schuchart | [1] | ||||
G11 | Towards multi-modal causability with Graph Neural Networks | Söchting | [1] | ||||
G12 | Explainability in Graph Neural Networks | Söchting | [1] | ||||
Signal processing | [1] | ||||||
SP1 | Location Tracking using Mobile Device Sensors | Rohde | [1], [2] | ||||
SP2 | Person tracking – lifted multicut | Leipnitz | [1] | ||||
SP3 | Person tracking – Graph similarity | Leipnitz | [1] | ||||
SP4 | Small Bird Recognition | Franke | [1] | ||||
Deep Learning in Physics | |||||||
PH1 | Physics Informed Deep Learning | Uhrich | [1] | ||||
PH2 | Deep Neural Networks motivated by Partial Differential Equations | Uhrich | [1] | ||||
Bio-Medical Applications | |||||||
BM1 | Viral Host Prediction by Deep Learning | Ewald | [1] | ||||
BM2 | Predicting Cancer Cells-lines’ Drug-Response Using a Probabilistic Graphical Model | Pogany | [1] | ||||
BM3 | Covid-19 Knowledge Graph | Christen | [1], [2] | ||||
BM4 | Interpretable Predictions of biomedical relationships via knowledge graphs | Christen | [1], [2] | ||||
BM5 | Classification and Visualization of Alzheimer’s Disease using Volumetric Convolutional Neural Network and Transfer Learning | Neumann | [1], [2] | ||||
BM6 | OM-Nets with Cross-task Guided Attention for Brain Tumor Segmentation | Neumann | [1],[2] | ||||
BM7 | Can Virtual Contrast Enhancement in Brain MRI Replace Gadolinium? | Martin | [1] | ||||
BM8 | 3D CNN for detection of Lesions in chest CT | Martin | [1] |